Explainable RL Policies by Distilling to Locally-Specialized Linear Policies with Voronoi State Partitioning

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Deproost, Senne, Steckelmacher, Dennis, Nowé, Ann
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866908659453263872
author Deproost, Senne
Steckelmacher, Dennis
Nowé, Ann
author_facet Deproost, Senne
Steckelmacher, Dennis
Nowé, Ann
contents Deep Reinforcement Learning is one of the state-of-the-art methods for producing near-optimal system controllers. However, deep RL algorithms train a deep neural network, that lacks transparency, which poses challenges when the controller has to meet regulations, or foster trust. To alleviate this, one could transfer the learned behaviour into a model that is human-readable by design using knowledge distilla- tion. Often this is done with a single model which mimics the original model on average but could struggle in more dynamic situations. A key challenge is that this simpler model should have the right balance be- tween flexibility and complexity or right balance between balance bias and accuracy. We propose a new model-agnostic method to divide the state space into regions where a simplified, human-understandable model can operate in. In this paper, we use Voronoi partitioning to find regions where linear models can achieve similar performance to the original con- troller. We evaluate our approach on a gridworld environment and a classic control task. We observe that our proposed distillation to locally- specialized linear models produces policies that are explainable and show that the distillation matches or even slightly outperforms the black-box policy they are distilled from.
format Preprint
id arxiv_https___arxiv_org_abs_2511_13322
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Explainable RL Policies by Distilling to Locally-Specialized Linear Policies with Voronoi State Partitioning
Deproost, Senne
Steckelmacher, Dennis
Nowé, Ann
Machine Learning
Artificial Intelligence
Deep Reinforcement Learning is one of the state-of-the-art methods for producing near-optimal system controllers. However, deep RL algorithms train a deep neural network, that lacks transparency, which poses challenges when the controller has to meet regulations, or foster trust. To alleviate this, one could transfer the learned behaviour into a model that is human-readable by design using knowledge distilla- tion. Often this is done with a single model which mimics the original model on average but could struggle in more dynamic situations. A key challenge is that this simpler model should have the right balance be- tween flexibility and complexity or right balance between balance bias and accuracy. We propose a new model-agnostic method to divide the state space into regions where a simplified, human-understandable model can operate in. In this paper, we use Voronoi partitioning to find regions where linear models can achieve similar performance to the original con- troller. We evaluate our approach on a gridworld environment and a classic control task. We observe that our proposed distillation to locally- specialized linear models produces policies that are explainable and show that the distillation matches or even slightly outperforms the black-box policy they are distilled from.
title Explainable RL Policies by Distilling to Locally-Specialized Linear Policies with Voronoi State Partitioning
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2511.13322